huggingface/candle · error

two elements have different len {m} {}

Error message

two elements have different len {m} {}

What it means

When building a 2-D tensor from Vec<&[S]>, all rows must have the same length. NdArray::shape compares each row's length to the first row's and bails on mismatch, since ragged data cannot form a rectangular Shape.

Source

Thrown at candle-core/src/device.rs:125

    fn shape(&self) -> Result<Shape> {
        Ok(Shape::from(self.len()))
    }

    fn to_cpu_storage(&self) -> CpuStorage {
        S::to_cpu_storage(self.as_slice())
    }
}

impl<S: WithDType> NdArray for Vec<&[S]> {
    fn shape(&self) -> Result<Shape> {
        if self.is_empty() {
            crate::bail!("empty array")
        }
        let n = self.len();
        let m = self[0].len();
        for v in self.iter() {
            if v.len() != m {
                crate::bail!("two elements have different len {m} {}", v.len())
            }
        }
        Ok(Shape::from((n, m)))
    }

    fn to_cpu_storage(&self) -> CpuStorage {
        let data = self.iter().copied().flatten().copied().collect::<Vec<_>>();
        S::to_cpu_storage_owned(data)
    }
}

impl<S: WithDType> NdArray for Vec<Vec<S>> {
    fn shape(&self) -> Result<Shape> {
        if self.is_empty() {
            crate::bail!("empty array")
        }
        let n = self.len();
        let m = self[0].len();

View on GitHub (pinned to d5fee525bf)

Solutions

1. Pad all rows to the same length before constructing the tensor.
2. Validate row lengths up front and report which row differs.
3. Build a 1-D tensor and reshape manually if the data is genuinely ragged.

Example fix

// before
let t = Tensor::new(vec![&[1.0, 2.0][..], &[3.0][..]], &dev)?;
// after
let t = Tensor::new(vec![&[1.0, 2.0][..], &[3.0, 0.0][..]], &dev)?;

let m = rows[0].len();
if rows.iter().any(|r| r.len() != m) {
    return Err(anyhow::anyhow!("rows must all have length {m}"));
}

When it happens

Trigger: Tensor::new / from_slice with Vec<&[S]> where rows have differing lengths (e.g. [[1,2],[3]]).

Common situations: Tokenized variable-length sequences passed without padding; CSV/parsed rows of uneven width.

Related errors

• two elements have different shapes {shape:?} {shape0:?}
• empty array
• backward not supported for non uniform upscaling factors
• in_channel mismatch between input ({c_in}) and kernel ({c_in
• in_channel {c_in} is not divisible by the number of groups

AI-assisted analysis of huggingface/candle@d5fee525bf (2026-09-02). Data as JSON: /api/errors/b68df835717be4ee. Report an issue: GitHub.